From: Jiri Olsa jolsa@kernel.org

commit 117211aa739a926e6555cfea883be84bee6f1695 upstream.

Pengfei Xu reported [1] Syzkaller/KASAN issue found in bpf_link_show_fdinfo.

The reason is missing BPF_LINK_TYPE invocation for uprobe multi link and for several other links, adding that.

[1] https://lore.kernel.org/bpf/ZXptoKRSLspnk2ie@xpf.sh.intel.com/

Fixes: 89ae89f53d20 ("bpf: Add multi uprobe link") Fixes: e420bed02507 ("bpf: Add fd-based tcx multi-prog infra with link support") Fixes: 84601d6ee68a ("bpf: add bpf_link support for BPF_NETFILTER programs") Fixes: 35dfaad7188c ("netkit, bpf: Add bpf programmable net device") Reported-by: Pengfei Xu pengfei.xu@intel.com Signed-off-by: Jiri Olsa jolsa@kernel.org Signed-off-by: Andrii Nakryiko andrii@kernel.org Tested-by: Pengfei Xu pengfei.xu@intel.com Acked-by: Hou Tao houtao1@huawei.com Link: https://lore.kernel.org/bpf/20231215230502.2769743-1-jolsa@kernel.org Cc: stable@vger.kernel.org # 6.6 Signed-off-by: Ignat Korchagin ignat@cloudflare.com --- Hi,

We have experienced a KASAN warning in production on a 6.6 kernel, similar to [1]. This backported patch was adjusted to apply onto 6.6 stable branch: the only change is dropping the BPF_LINK_TYPE(BPF_LINK_TYPE_NETKIT, netkit) definition from the header as netkit was only introduced in 6.7 and 6.7 has the backport already.

I was not able to run the syzkaller reproducer from [1], but we have not seen the KASAN warning in production since applying this patch internally.

Regards, Ignat

include/linux/bpf_types.h | 3 +++ 1 file changed, 3 insertions(+)

diff --git a/include/linux/bpf_types.h b/include/linux/bpf_types.h index fc0d6f32c687..dfaae3e3ec15 100644 --- a/include/linux/bpf_types.h +++ b/include/linux/bpf_types.h @@ -142,9 +142,12 @@ BPF_LINK_TYPE(BPF_LINK_TYPE_ITER, iter) #ifdef CONFIG_NET BPF_LINK_TYPE(BPF_LINK_TYPE_NETNS, netns) BPF_LINK_TYPE(BPF_LINK_TYPE_XDP, xdp) +BPF_LINK_TYPE(BPF_LINK_TYPE_NETFILTER, netfilter) +BPF_LINK_TYPE(BPF_LINK_TYPE_TCX, tcx) #endif #ifdef CONFIG_PERF_EVENTS BPF_LINK_TYPE(BPF_LINK_TYPE_PERF_EVENT, perf) #endif BPF_LINK_TYPE(BPF_LINK_TYPE_KPROBE_MULTI, kprobe_multi) BPF_LINK_TYPE(BPF_LINK_TYPE_STRUCT_OPS, struct_ops) +BPF_LINK_TYPE(BPF_LINK_TYPE_UPROBE_MULTI, uprobe_multi)